arXiv subjects

Taiquan Wei

Publications and source records attributed to Taiquan Wei.

6 recordsLinked to original sources

arXiv2026

MOCAP: Wafer-Scale-Chip-Oriented Memory-Orchestrated Chunked Pipelining Framework for Prefill-Only LLM Inference

Large language models (LLMs) are increasingly used in prefill-only workloads, where end-to-end latency is dominated by the prefill phase. For long-context prefill, communication overhead grows with sequence length and quickly becomes a bottleneck on conventional GPU systems, making wafer-scale chips (WSCs) a promising substrate due to their high communication bandwidth and large aggregate compute and memory capacity. A natural way to accelerate prefill is to partition a long input sequence into multiple chunks and execute them in a finer-grained pipeline across devices. However, directly applying this idea to long-context prefill on WSCs remains challenging. First, causal dependency across chunks causes KV cache to accumulate unevenly across pipeline stages, creating severe memory imbalance and limiting the feasible sequence length. Second, later chunks require more attention computation because each chunk depends on preceding chunks, leading to chunk-level latency imbalance. To address these challenges, we present MOCAP, a memory-orchestrated chunked pipelining framework for prefill-only LLM inference on WSCs. MOCAP introduces Memory-Balanced KV Reallocation (MBKR) to alleviate memory imbalance by redistributing KV cache across pipeline stages, thereby extending the feasible sequence length. It further incorporates Latency-Balanced Chunk Partitioning (LBCP) to balance chunk execution cost under both attention-cost growth and KV reallocation overhead, improving pipeline efficiency. Experimental results show that, compared with GPipe, MOCAP achieves 76.4\% lower end-to-end latency and 3.24$\times$ higher throughput on average. MOCAP also extends the maximum supported sequence length by up to 1.31$\times$ compared with Terapipe.

cs.AR↗

arXiv2025

Designing Spatial Architectures for Sparse Attention: STAR Accelerator via Cross-Stage Tiling

Large language models (LLMs) rely on self-attention for contextual understanding, demanding high-throughput inference and large-scale token parallelism (LTPP). Existing dynamic sparsity accelerators falter under LTPP scenarios due to stage-isolated optimizations. Revisiting the end-to-end sparsity acceleration flow, we identify an overlooked opportunity: cross-stage coordination can substantially reduce redundant computation and memory access. We propose STAR, a cross-stage compute- and memory-efficient algorithm-hardware co-design tailored for Transformer inference under LTPP. STAR introduces a leading-zero-based sparsity prediction using log-domain add-only operations to minimize prediction overhead. It further employs distributed sorting and a sorted updating FlashAttention mechanism, guided by a coordinated tiling strategy that enables fine-grained stage interaction for improved memory efficiency and latency. These optimizations are supported by a dedicated STAR accelerator architecture, achieving up to 9.2$\times$ speedup and 71.2$\times$ energy efficiency over A100, and surpassing SOTA accelerators by up to 16.1$\times$ energy and 27.1$\times$ area efficiency gains. Further, we deploy STAR onto a multi-core spatial architecture, optimizing dataflow and execution orchestration for ultra-long sequence processing. Architectural evaluation shows that, compared to the baseline design, Spatial-STAR achieves a 20.1$\times$ throughput improvement.

cs.AR↗

arXiv2025

TEMP: A Memory Efficient Physical-aware Tensor Partition-Mapping Framework on Wafer-scale Chips

Large language models (LLMs) demand significant memory and computation resources. Wafer-scale chips (WSCs) provide high computation power and die-to-die (D2D) bandwidth but face a unique trade-off between on-chip memory and compute resources due to limited wafer area. Therefore, tensor parallelism strategies for wafer should leverage communication advantages while maintaining memory efficiency to maximize WSC performance. However, existing approaches fail to address these challenges. To address these challenges, we propose the tensor stream partition paradigm (TSPP), which reveals an opportunity to leverage WSCs' abundant communication bandwidth to alleviate stringent on-chip memory constraints. However, the 2D mesh topology of WSCs lacks long-distance and flexible interconnects, leading to three challenges: 1) severe tail latency, 2) prohibitive D2D traffic contention, and 3) intractable search time for optimal design. We present TEMP, a framework for LLM training on WSCs that leverages topology-aware tensor-stream partition, traffic-conscious mapping, and dual-level wafer solving to overcome hardware constraints and parallelism challenges. These integrated approaches optimize memory efficiency and throughput, unlocking TSPP's full potential on WSCs. Evaluations show TEMP achieves 1.7x average throughput improvement over state-of-the-art LLM training systems across various models.

cs.AR↗

arXiv2025

WATOS: Efficient LLM Training Strategies and Architecture Co-exploration for Wafer-scale Chip

Training large language models (LLMs) imposes extreme demands on computation, memory capacity, and interconnect bandwidth, driven by their ever-increasing parameter scales and intensive data movement. Wafer-scale integration offers a promising solution by densely integrating multiple single-die chips with high-speed die-to-die (D2D) interconnects. However, the limited wafer area necessitates trade-offs among compute, memory, and communication resources. Fully harnessing the potential of wafer-scale integration while mitigating its architectural constraints is essential for maximizing LLM training performance. This imposes significant challenges for the co-optimization of architecture and training strategies. Unfortunately, existing approaches all fall short in addressing these challenges. To bridge the gap, we propose WATOS, a co-exploration framework for LLM training strategy and wafer-scale architecture. We first define a highly configurable hardware template designed to explore optimal architectural parameters for wafer-scale chips. Based on it, we capitalize on the high D2D bandwidth and fine-grained operation advantages inherent to wafer-scale chips to explore optimal parallelism and resource allocation strategies, effectively addressing the memory underutilization issues during LLM training. Compared to the state-of-the-art (SOTA) LLM training framework Megatron and Cerebras' weight streaming wafer training strategy, WATOS can achieve an average overall throughput improvement of 2.74x and 1.53x across various LLM models, respectively. In addition, we leverage WATOS to reveal intriguing insights about wafer-scale architecture design with the training of LLM workloads.

eess.SP↗

arXiv2025

MoEntwine: Unleashing the Potential of Wafer-scale Chips for Large-scale Expert Parallel Inference

As large language models (LLMs) continue to scale up, mixture-of-experts (MoE) has become a common technology in SOTA models. MoE models rely on expert parallelism (EP) to alleviate memory bottleneck, which introduces all-to-all communication to dispatch and combine tokens across devices. However, in widely-adopted GPU clusters, high-overhead cross-node communication makes all-to-all expensive, hindering the adoption of EP. Recently, wafer-scale chips (WSCs) have emerged as a platform integrating numerous devices on a wafer-sized interposer. WSCs provide a unified high-performance network connecting all devices, presenting a promising potential for hosting MoE models. Yet, their network is restricted to a mesh topology, causing imbalanced communication pressure and performance loss. Moreover, the lack of on-wafer disk leads to high-overhead expert migration on the critical path. To fully unleash this potential, we first propose Entwined Ring Mapping (ER-Mapping), which co-designs the mapping of attention and MoE layers to balance communication pressure and achieve better performance. We find that under ER-Mapping, the distribution of cold and hot links in the attention and MoE layers is complementary. Therefore, to hide the migration overhead, we propose the Non-invasive Balancer (NI-Balancer), which splits a complete expert migration into multiple steps and alternately utilizes the cold links of both layers. Evaluation shows ER-Mapping achieves communication reduction up to 62%. NI-Balancer further delivers 54% and 22% improvements in MoE computation and communication, respectively. Compared with the SOTA NVL72 supernode, the WSC platform delivers an average 39% higher per-device MoE performance owing to its scalability to larger EP.

cs.DC↗

arXiv2025

MCBP: A Memory-Compute Efficient LLM Inference Accelerator Leveraging Bit-Slice-enabled Sparsity and Repetitiveness

Large language models (LLMs) face significant inference latency due to inefficiencies in GEMM operations, weight access, and KV cache access, especially in real-time scenarios. This highlights the need for a versatile compute-memory efficient accelerator. Unfortunately, existing Transformer accelerators struggle to address both aspects simultaneously, as they focus on value-level processing, missing fine-grained opportunities to optimize computation and memory collaboratively. This paper introduces MCBP, a bit-grained compute-memory efficient algorithm-hardware co-design that leverages bit-slice (BS) enabled repetitiveness and sparsity to accelerate LLM inference. MCBP features three key innovations: 1) BS-repetitiveness-enabled computation reduction (BRCR), which eliminates redundant GEMM computations via leveraging redundancy hidden among BS vectors; 2) BS-sparsity-enabled two-state coding (BSTC), which reduces weight access via exploiting significant sparsity in high-order bit-slice weight; 3) Bit-grained progressive prediction (BGPP), which reduces KV cache access by leveraging early-termination-based bit-grained prediction. These techniques, supported by custom accelerator designs, effectively alleviate the burden in GEMM, weight access, and KV cache access. Extensive experiments on 26 benchmarks show that MCBP achieves 9.43x speed up and 31.1x higher energy efficiency than Nvidia A100 GPU. Compared to SOTA Transformer accelerators, MCBP achieves 35x, 5.2x and 3.2x energy saving than Spatten, FACT and SOFA, respectively.

cs.AR↗